A latent class analysis of parental bipolar disorder: examining associations with offspring psychopathology

Abstract

Bipolar disorder (BD) is highly heterogeneous, and course variations are associated with patient outcomes. This diagnostic complexity challenges identification of patients in greatest need of intervention. Additionally, course variations have implications for offspring risk. First, latent class analysis (LCA) categorized parents with BD based on salient illness characteristics: BD type, onset age, polarity of index episode, pole of majority of episodes, rapid cycling, psychosis, anxiety comorbidity, and substance dependence. Fit indices favored three parental classes with some substantively meaningful patterns. Two classes, labeled “Earlier-Onset Bipolar-I” (EO-I) and “Earlier-Onset Bipolar-II” (EO-II), comprised parents who had a mean onset age in mid-adolescence, with EO-I primarily BD-I parents and EO-II entirely BD-II parents. The third class, labeled “Later-Onset BD” (LO) had an average onset age in adulthood. Classes also varied on probability of anxiety comorbidity, substance dependence, psychosis, rapid cycling, and pole of majority of episodes. Second, we examined rates of disorders in offspring (ages 4–33, Mage=13.46) based on parental latent class membership. Differences emerged for offspring anxiety disorders only such that offspring of EO-I and EO-II parents had higher rates, compared to offspring of LO parents, particularly for daughters. Findings may enhance understanding of BD and its nosology